[Lustre-discuss] LustreError

Mohr Jr, Richard Frank (Rick Mohr) rmohr at utk.edu
Thu Jun 19 10:05:27 PDT 2014


( I have re-added the lustre-discuss mailing list to the reply.)

I am not familiar with that error message.  A quick google turned up a couple of links that may be helpful to you:

http://lists.lustre.org/pipermail/lustre-discuss/2010-September/014035.html
https://jira.hpdd.intel.com/browse/LU-275

Someone else on the mailing list might have more knowledge about this.

-- 
Rick Mohr
Senior HPC System Administrator
National Institute for Computational Sciences
http://www.nics.tennessee.edu


On Jun 19, 2014, at 12:33 PM, "<>" <curiojustus at gmail.com>
 wrote:

> Dear Mohr,
> 
> Thanks for your reply.
> A RAID card was found faulty, and it was changed.
> But now the error message has change to:
> kernel: blk_rq_check_limits: over max size limit.
> my understanding is that it is because of max_hw_sectors_kb.
> I had forgot to mention that we are trying to use a loop device image as mds server which:
> cat /sys/block/loop0/queue/max_hw_sectors_kb
> 127
>  
> Could it be the cause? Thanks
> 
> 
> 
> 
> 
> 
> 
> 2014-06-16 19:02 GMT+02:00 Mohr Jr, Richard Frank (Rick Mohr) <rmohr at utk.edu>:
> 
> On Jun 5, 2014, at 11:48 AM, "<>" <curiojustus at gmail.com>
>  wrote:
> 
> > [user at disk-server]$ lctl dl
> <snip>
> >  12 ST obdfilter lustre-OST0000 lustre-OST0000_UUID 5
> 
> One of your OSTs appears to be down which would explain why "lfs df" was hanging.  Have you been able to troubleshoot this problem to determine the cause?  Do you have a hardware failure?
> 
> --
> Rick Mohr
> Senior HPC System Administrator
> National Institute for Computational Sciences
> http://www.nics.tennessee.edu
> 
> 





More information about the lustre-discuss mailing list